Cholecystokin (CCK), a 33-residue peptide, plays a major role in the gastrointestinal system (gall bladder contraction, amylase secretion, gut motility, food intake) as well as in the CNS as a neurotransmitter and/or neuromodulator. The CCK system, besides being of considerable therapeutic interest, offers a unique opportunity to decipher the structural basis of agonist action at its two receptor subtypes. This is due to the availability of a diverse set of peptide analogs, both agonist and antagonists, as well as non-peptide analogs, all antagonists. Models of the receptor-bound conformations of CCK at the two receptors provides a good interpretation of much of the current literature on peptide analogs. We propose to confirm and refine these models by synthesis of conformationally constrained analogs. The refined models will allow the alignment of the non-peptide antagonists and the :proposal of modifications which should convert them to agonists. This would demon- strate that the models are valid and that we understand the structural basis of the differences between agonists and antagonists. The ultimate goal of this proposal is a predictive, quantitative model for the two CCK receptor subtypes which will allow the design and synthesis of novel therapeutic agents.